pivotal / cf-onboarding Goto Github PK
View Code? Open in Web Editor NEWIterating on Onboarding Week with lightning speed and cunning
License: MIT License
Iterating on Onboarding Week with lightning speed and cunning
License: MIT License
Have an epic that would cover container networking.
A feature of security groups that we didn't use in the security group stories was creating and binding a new security group. This seems to involve writing a json file and then either running a cf command or directly making requests against the cc api (?!).
docs: https://docs.cloudfoundry.org/concepts/asg.html#procedures
@SimonParker
A question we weren't able to answer for ourselves:
Why does cf create-route exist? Wouldn't it be enough to only have cf map-route? It sounds like creating but not mapping a route leaves it useless, and the process of mapping a route is exactly the same whether or not you created the route previously.
re: inter- and intra-component routing in CF
Both bbl and bosh accept ops files. Add a story that involves creating an ops-file (rather than just passing one in).
Explain how to create a load balancer for Concourse without bbl that will work with deployment instructions so that the CF lb doesn't have to be torn down.
Have a ready to go Concourse env so one could have a track that is dedicated to only setting up pipelines and not the concourse env.
... so that the instructions are less confusing
In the Nameservers section, select the Use custom nameservers (enter below) radio button.
Otherwise, your Namerserver will remain the default.
This has been observed once and may be a user configuration error.
If you create your redis db with RedisLabs, you need to specify your username when creating the cf service via cf cups
...e.g.
cf cups redis -p "host, port, username, password"
This username will be whatever email address you used to sign up for the service with.
This should be more clear. In the .yml, you should edit the workers
line under the compilation
section.
The stories relating to bosh deploy
should clarify what -o
does and what the files in the operations
folder do.
The wording makes the use of the operation folder feel like it must be applied, but in reality the files under operations
are just options one can activate, and at any time after the creation of a bosh deployed cluster as long as one executes the bosh deploy
.
Most of the epics and stories in this backlog are focused on the open source cloud foundry and bosh, however many teams work on PCF and PCF tiles. It would awesome if there was an epic which went through the opsman and ERT lifecycle.
In addition to this, it would be great to also install a service tile and go through the interaction between PCF and a service.
cc @kkallday
Describe what traffic we expect to block
story: https://www.pivotaltracker.com/story/show/145951205
The password is not the one used when creating your account. The database defaults to not having a password, and you can configure it to require one. I don't think it needs the username either, just the password (if you set it up).
Most CF software is intended to be consumed as a BOSH release, so it would be really helpful to have an epic which goes through the entire process.
And what is creating so many of them? And why are they not being destroyed when whatever is using them is destroyed?
The story https://www.pivotaltracker.com/story/show/142952109 contains instructions to clone the repo, which is linked earlier in the story. This is a little confusing as there is not really an indication that the link goes to a repo, or an association between the word "repo" and the link.
Suggestion:
Make the word "repo" the link to the repository.
http://concourse.ci/vagrant.html gives me a 404 error :(
https://www.pivotaltracker.com/story/show/150192732
It isn't immediately clear from the story that the user must set the path for -keyout and -out. It would be clearer if the intended command were expressed in a way similar to the below:
openssl req -x509 -newkey rsa:2048 -keyout CHOOSE-PRIVATE-KEY-FILE-NAME.pem -out CHOOSE-CERT-FILE-NAME.pem -nodes
The link: https://docs.cloudfoundry.org/adminguide/uaa-user-management.html leads to "page not found". The link is referenced in both the "How?" and "Resources" sections.
not super clear where to get app-instance-index from when SSH'ing into our newly deployed app. Realized that it's the zero index of the number of instances you specified in the manifest.yml, but took us a second... (https://www.pivotaltracker.com/story/show/142969525)
Be explicit that user must run bosh update-cloud-config
to update cloud config
This intro on docker's website really helped us understand how to build a docker image better than the article linked to in the story (https://www.pivotaltracker.com/story/show/143145055). We ended up installing Docker for Mac and using it that way.
In this story, we set up iam service accounts. It would be helpful to have some explanation of what iam and service accounts are.
@SimonParker
Some confusion on this story. Text reads:
"Then, follow these steps to set up your GCP service account, replacing ACCOUNT_NAME with the name you just chose and PROJECT_ID with your assigned project's name"
We're not finding any reference to choosing a name for this account. Presumably, we can make up anything we want? The fact that we pull it from a previous step suggests it needs to correspond with something, but we don't know what.
Pivotal uses GITBOT to synchronize Github issues and pull requests with Pivotal Tracker.
Please add your new repo to the GITBOT config-production.yml
in the Gitbot configuration repo.
If you don't have access you can send an ask ticket to the CF admins. We prefer teams to submit their changes via a pull request.
Steps:
config-production.yml
fileIf there are any questions, please reach out to [email protected].
Let people know to ignore the "Update your DNS records to point to your GCP load balancer" step in lbs instructions.
..instead of just being copy-pastable
story: https://www.pivotaltracker.com/story/show/145951219
Claims that the security group should be all_open
on gcp, but we see dns
and public_networks
. They still seem to be named reasonably -- dns
allows port 53 on tcp or udp.
From the story:
#### Expected Result
If you've deployed a full Cloud Foundry on GCP you should have only one security group, applied to both staging and running apps: `all_open`.
If you're working with PCF Dev, you should see three security groups, one of which is named `all_pcfdev`.
For either circumstance if you run `cf security-groups <SECURITY-GROUP NAME>` you'll see that they do exactly what it sounds like they do—leave everything open by allowing containers to access all IPs on any port.
Because of the `all_open` (or `all_pcfdev`) security group any other group would be redundant.
The output we see:
→ cf security-groups
Getting security groups as onboarding
OK
Name Organization Space
#0 public_networks
#1 dns
→ cf security-group public_networks
Getting info for security group public_networks as onboarding
OK
Name public_networks
Rules
[
{
"destination": "0.0.0.0-9.255.255.255",
"protocol": "all"
},
{
"destination": "11.0.0.0-169.253.255.255",
"protocol": "all"
},
{
"destination": "169.255.0.0-172.15.255.255",
"protocol": "all"
},
{
"destination": "172.32.0.0-192.167.255.255",
"protocol": "all"
},
{
"destination": "192.169.0.0-255.255.255.255",
"protocol": "all"
}
]
No spaces assigned
→ cf security-group dns
Getting info for security group dns as onboarding
OK
Name dns
Rules
[
{
"destination": "0.0.0.0/0",
"ports": "53",
"protocol": "tcp"
},
{
"destination": "0.0.0.0/0",
"ports": "53",
"protocol": "udp"
}
]
No spaces assigned
@SimonParker
This story stumped us; our CPU quota is exactly 24, so per instructions, we need to decrease the number of compilation workers.
In the cloud config generated from bosh, we have compilation.workers: 6. We didn't know if that was too high, or what number we needed to decrease it to. 5? 4?
We decided to plow ahead and when we hit the story requiring bosh -d cf deploy, see if we got the CPU quota error, and fix it then. (We did wonder whether the reason this fix is done preemptively was because hitting the CPU quota is somehow expensive?)
In this story https://www.pivotaltracker.com/story/show/150296148
Some of us got hung up because we were logged in as User rather than Admin. It would help to have an instruction to login as Admin in this story.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.